What are we really studying? Exploring strategic decision-making in the small businesses literature using natural language processing models
Revista Ibero-Americana de Estratégia · 12.05.2026 · 25 (1) · e28792
DOI: 10.5585/2026.28792 ↗Kurz gefasst
Fachartikel in der Zeitschrift Revista Ibero-Americana de Estratégia, online erschienen am 12.05.2026. Verfasst von Alan Souza Lima und 1 weiteren Autor:innen. Die Arbeit nennt 94 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-SA 4.0).
Abstract (Originalsprache)
Study objective: This paper critically examines how strategic decision-making has been conceptualized in small business research, identifying patterns of theoretical fragmentation and conceptual reification. The aim is to clarify the field’s current state and propose directions for a more coherent understanding of strategic management in small firms. Methodology/approach: We conducted a systematic literature review of 71 peer-reviewed articles indexed in the Web of Science (WoS) database. To analyze this corpus, we employed a triangulated natural language processing (NLP) framework combining three approaches: Iramuteq (Reinert Method) to uncover explicit structural patterns in the literature; Latent Dirichlet Allocation (LDA) to identify probabilistic and cross-cutting thematic clusters; and BERTopic to detect latent semantic relationships and emerging conceptual proximities. This multi-method approach enables a comprehensive mapping of both visible and implicit structures shaping the field. Originality/Relevance: By integrating computational text analysis into a systematic review, this study moves beyond descriptive synthesis to show how the field’s conceptual architecture has evolved and fractured over time. It exposes the rhetorical use of “strategy” without definitional grounding, showing how theoretical borrowing from unrelated domains has diluted conceptual coherence. Main results: Findings show that research on strategic decision-making in small businesses is predominantly owner-centered, methodologically diverse but weakly integrated, and rarely grounded in strategic management theory. The combination of Iramuteq, LDA, and BERTopic points to distinct but convergent evidence of fragmentation, both lexical and semantic, across decades of publications. Theoretical/Methodological contributions: The study introduces a multi-layered analytical framework for reconstructing dispersed research fields. The triangulated NLP approach allows for simultaneous analysis of explicit discourse, thematic configurations, and underlying semantic relations, providing a methodological model that extends the scope of systematic literature reviews. Social/Management contributions: Clarifying how small business research frames strategic decision-making can guide scholars and practitioners toward more consistent theoretical models and structured decision processes. Such alignment can improve managerial effectiveness, inclusiveness, and resilience in small enterprises.
Zitieren
APA
Lima, A. S., & Martins, F. S. (2026). What are we really studying? Exploring strategic decision-making in the small businesses literature using natural language processing models. Revista Ibero-Americana de Estratégia, 25(1), e28792. https://doi.org/10.5585/2026.28792
Vancouver
Lima AS, Martins FS. What are we really studying? Exploring strategic decision-making in the small businesses literature using natural language processing models. Revista Ibero-Americana de Estratégia. 2026;25(1):e28792. doi:10.5585/2026.28792
BibTeX
@article{Lima2026,
author = {Lima, Alan Souza and Martins, Fellipe Silva},
title = {What are we really studying? Exploring strategic decision-making in the small businesses literature using natural language processing models},
journal = {Revista Ibero-Americana de Estratégia},
year = {2026},
volume = {25},
number = {1},
pages = {e28792},
doi = {10.5585/2026.28792}
}
EUVIMED